CN113160183A - Hyperspectral data processing method, device and medium - Google Patents
Hyperspectral data processing method, device and medium Download PDFInfo
- Publication number
- CN113160183A CN113160183A CN202110453331.0A CN202110453331A CN113160183A CN 113160183 A CN113160183 A CN 113160183A CN 202110453331 A CN202110453331 A CN 202110453331A CN 113160183 A CN113160183 A CN 113160183A
- Authority
- CN
- China
- Prior art keywords
- principal component
- hyperspectral image
- hyperspectral
- region
- detected
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 8
- 238000003062 neural network model Methods 0.000 claims abstract description 37
- 238000000513 principal component analysis Methods 0.000 claims abstract description 29
- 238000012847 principal component analysis method Methods 0.000 claims abstract description 23
- 238000012545 processing Methods 0.000 claims abstract description 22
- 238000012549 training Methods 0.000 claims abstract description 20
- 230000003595 spectral effect Effects 0.000 claims abstract description 17
- 238000000034 method Methods 0.000 claims description 38
- 238000001228 spectrum Methods 0.000 claims description 21
- 238000001514 detection method Methods 0.000 claims description 13
- 238000002310 reflectometry Methods 0.000 claims description 6
- 238000004519 manufacturing process Methods 0.000 claims description 4
- 238000012937 correction Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 10
- 238000004590 computer program Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000000701 chemical imaging Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000013136 deep learning model Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 241000607479 Yersinia pestis Species 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000003795 chemical substances by application Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
- G06T2207/10036—Multispectral image; Hyperspectral image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20092—Interactive image processing based on input by user
- G06T2207/20104—Interactive definition of region of interest [ROI]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Quality & Reliability (AREA)
- Bioinformatics & Computational Biology (AREA)
- Geometry (AREA)
- Image Analysis (AREA)
Abstract
The application discloses a hyperspectral data processing method, hyperspectral data processing equipment and a hyperspectral data processing medium. And performing principal component analysis on each hyperspectral image in the hyperspectral image set according to a principal component analysis method to obtain first principal component information with a first preset number. And determining the corresponding characteristic wave band according to the weight coefficient curve in the first principal component system information. And performing principal component analysis on the characteristic wave band according to a principal component analysis method to determine second principal component information with a second preset number. And taking the spectral characteristics corresponding to the second principal component information and the corresponding hyperspectral image as sample data. And training and detecting the neural network model according to the sample data to obtain the trained neural network model corresponding to the object to be detected.
Description
Technical Field
The present application relates to the field of hyperspectral technology, and in particular, to a hyperspectral data processing method, device, and medium.
Background
With the continuous development of science and technology, the hyperspectral imaging technology is widely applied. The application range of hyperspectral imaging extends to a plurality of fields of chemistry, physics, biology, medicine and the like. At present, hyperspectral imaging has wide and profound application prospects in the aspects of land utilization, crop growth, classification, pest and disease detection, marine water color measurement, urban planning, oil exploration, earth core landform, military target identification and the like.
The hyperspectral image integrates image information and spectral information, and has the characteristics of multiple wave bands, rich intrinsic information, high spectral resolution and the like, the information contained in the hyperspectral image can reflect external characteristics such as the size, the shape, the volume and the like of a sample, and the characteristics determine the unique advantages of the hyperspectral image in tasks needing internal and external characteristic detection. However, the rich information contained in the hyperspectral image makes the hyperspectral image occupy a large amount of data when being expressed, which also makes the deep learning model complex, the model is slow in the training process and the testing process, and a large amount of training samples are needed.
Based on the above, a technical scheme for processing the hyperspectral data is provided, the hyperspectral images are processed, so that the model training and testing can be completed by fewer processed hyperspectral images of the deep learning model, and the technical problem of continuous solution is solved.
Disclosure of Invention
The embodiment of the specification provides hyperspectral data processing, hyperspectral data processing equipment and a hyperspectral data processing medium, and is used for solving the following technical problems in the prior art: when a neural network model related to a hyperspectral image is constructed, due to the particularity of the hyperspectral image, a large amount of sample data is needed when the neural network model is constructed, and the training and testing speeds are slow.
A method of hyperspectral data processing, the method comprising:
acquiring a hyperspectral image set of an object to be detected;
performing principal component analysis on each hyperspectral image in the hyperspectral image set according to a principal component analysis method to obtain first principal component information of a first preset number;
determining a corresponding characteristic waveband according to a weight coefficient curve in the first principal component system information;
performing principal component analysis on the characteristic wave band according to a principal component analysis method, and determining second principal component information with a second preset number;
taking the spectral characteristics corresponding to the second principal component information and the corresponding hyperspectral image as sample data;
and training and detecting a neural network model according to the sample data to obtain a trained neural network model corresponding to the object to be detected.
In one possible implementation manner, before acquiring the hyperspectral image set of the object to be detected, the method further includes:
determining an object to be detected for constructing a neural network model based on input information of a user;
acquiring an original hyperspectral image of the object to be detected from a prestored database based on the object to be detected and a preset incidence relation;
carrying out target detection on each original hyperspectral image to obtain a corresponding region of interest;
segmenting the corresponding original hyperspectral image according to the region of interest to obtain a region of interest image;
and the images of the interested areas form a hyperspectral image set of the object to be detected.
In a possible implementation manner, after obtaining the trained neural network model corresponding to the object to be detected, the method further includes:
receiving an initial hyperspectral image sent by hyperspectral image acquisition equipment;
carrying out normalization processing on the initial hyperspectral image to obtain a hyperspectral image to be detected;
and carrying out image recognition on the hyperspectral image to be detected based on the trained neural network model, and determining whether the initial hyperspectral image comprises the object to be detected.
In a possible implementation manner, performing target detection on each of the original hyperspectral images to obtain a corresponding region of interest specifically includes:
determining seed pixels of each original hyperspectral image;
based on a region growing algorithm and the seed pixels, carrying out image recognition on each original hyperspectral image to determine target contour information in each original hyperspectral image;
and dividing the corresponding hyperspectral image according to the target contour information to obtain a corresponding region-of-interest image.
In a possible implementation manner, the determining a seed pixel of each original hyperspectral image specifically includes:
determining a local window with a preset size, calculating the spectral similarity of pixels in the local window, and performing AP clustering to obtain a local window pixel clustering result;
obtaining a local window clustering mark graph according to the clustering result;
searching according to a preset square grid in the clustering label graph;
and when the gray values of all pixels in the preset square are equal, taking the central pixel of the square as a seed pixel.
In a possible implementation manner, based on a region growing algorithm and the seed pixel, performing image recognition on each of the original hyperspectral images to determine target contour information in each of the original hyperspectral images, specifically including:
based on a region production algorithm and the seed pixels, carrying out image recognition on the original hyperspectral image to obtain at least one growing region;
determining whether each of the growth regions is a connected region;
and calculating the area of the connected region, and determining the target region according to the area of the connected region to obtain corresponding target contour information.
In a possible implementation manner, performing target detection on each of the original hyperspectral images to obtain a corresponding region of interest specifically includes:
determining an initial seed pixel of a hyperspectral image, wherein a spectrum related to the seed is an initial seed spectrum;
calculating the spectrum difference between the pixel and the seed spectrum, and judging whether the pixel is grown according to the spectrum difference;
after the pixels are grown, updating the initial seed spectrum to the spectrum mean value of all grown pixels until the region growth is finished, and segmenting the background region;
and determining a new initial seed pixel, re-growing the region, and segmenting the background region until the new initial seed pixel cannot be determined so as to obtain a corresponding region of interest.
In one possible implementation, before obtaining the corresponding region of interest, the method further includes:
performing baseline correction on the original hyperspectral image; and
calculating a calibration coefficient based on the reflectivity of the object to be detected and the reflectivity of the standard reference plate;
and carrying out radiometric calibration on the corrected hyperspectral image according to the calibration coefficient to obtain the hyperspectral image after auxiliary calibration.
A hyperspectral data processing apparatus, the apparatus comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor;
wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to:
acquiring a hyperspectral image set of an object to be detected;
performing principal component analysis on each hyperspectral image in the hyperspectral image set according to a principal component analysis method to obtain first principal component information of a first preset number;
determining a corresponding characteristic waveband according to a weight coefficient curve in the first principal component system information;
performing principal component analysis on the characteristic wave band according to a principal component analysis method, and determining second principal component information with a second preset number;
taking the spectral characteristics corresponding to the second principal component information and the corresponding hyperspectral image as sample data;
and training and detecting a neural network model according to the sample data to obtain a trained neural network model corresponding to the object to be detected.
A non-transitory computer storage medium of hyperspectral data processing storing computer-executable instructions arranged to:
acquiring a hyperspectral image set of an object to be detected;
performing principal component analysis on each hyperspectral image in the hyperspectral image set according to a principal component analysis method to obtain first principal component information of a first preset number;
determining a corresponding characteristic waveband according to a weight coefficient curve in the first principal component system information;
performing principal component analysis on the characteristic wave band according to a principal component analysis method, and determining second principal component information with a second preset number;
taking the spectral characteristics corresponding to the second principal component information and the corresponding hyperspectral image as sample data;
and training and detecting a neural network model according to the sample data to obtain a trained neural network model corresponding to the object to be detected.
The embodiment of the specification adopts at least one technical scheme which can achieve the following beneficial effects: and performing secondary principal component analysis through two times of principal component analysis to obtain secondary principal component analysis, and performing neural network model training based on the spectral characteristics corresponding to the secondary principal component information and the corresponding hyperspectral image as sample data, so that the accuracy of the neural network model training is ensured, the sample data can be reduced, and the training and detection speed is improved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:
FIG. 1 is a flowchart of a hyperspectral data processing method provided by an embodiment of the present specification;
FIG. 2 is another flowchart of a hyperspectral data processing method provided by an embodiment of the specification;
FIG. 3 is another flowchart of a hyperspectral data processing method provided by an embodiment of the specification;
fig. 4 is a schematic structural diagram of a hyperspectral data processing device according to an embodiment of the specification.
Detailed Description
In order to make the objects, technical solutions and advantages of the present disclosure more apparent, the technical solutions of the present disclosure will be clearly and completely described below with reference to the specific embodiments of the present disclosure and the accompanying drawings. It should be apparent that the described embodiments are only some of the embodiments of the present application, and not all of the embodiments. All other embodiments obtained by a person skilled in the art without making any inventive step based on the embodiments in the description belong to the protection scope of the present application.
The technical solutions provided by the embodiments of the present application are described in detail below with reference to the accompanying drawings.
Fig. 1 is a flowchart of hyperspectral data processing provided by an embodiment of the specification. As shown in fig. 1, the hyperspectral food detection method provided by the application may include the following steps:
s101, the server obtains a hyperspectral image set of the object to be detected.
S102, the server performs principal component analysis on each hyperspectral image in the hyperspectral image set according to a principal component analysis method to obtain first principal component information with a first preset number.
And the server performs principal component analysis on the waveband information in the hyperspectral image by a principal component analysis method, and reserves at least the first N principal component information as first principal component information according to the corresponding preset number N. The reserved first N principal component information basically comprises most of available waveband information, so that the execution time is simplified, and the identification efficiency can be improved. Moreover, the image is processed by a principal component analysis method, redundant information among frequency bands is removed, and multi-band image information is compressed into a small amount, which is more effective than the original frequency band.
S103, the server determines a corresponding characteristic waveband according to the weight coefficient curve in the first main component system information.
In embodiments of the present application, the server may determine, according to the feature connection AG corresponding to each extracted first principal component information, a corresponding weight coefficient curve, and determine a feature band, where the feature band is at least a band in which weight values in two pieces of first principal component information are peak values or valley values at the same time.
And S104, the server performs principal component analysis on the characteristic wave band according to the principal component analysis method, and determines second principal component information with a second preset number.
Specifically, the server performs principal component analysis again on the characteristic band by a principal component analysis method to determine second principal component information. For example, the first principal component information retains the first 6 principal component information, and then the second principal component analysis is performed based on the characteristic band, retaining at least the first 3 principal component information as the second principal component information.
In the embodiment of the application, most of the first principal component information contains relatively more strong information, and most of the strong information reflects the principal information to be detected, so that no relevant emphasis exists. Therefore, in the embodiment of the application, the first principal component information is not selected for identification and analysis, but secondary principal component analysis is performed through the determined characteristic wave band to obtain the principal component information related to the construction of the neural network model, so that the identification efficiency is improved, and the influence is reduced.
And S105, the server takes the spectral characteristics corresponding to the second principal component information and the corresponding hyperspectral image as sample data.
And S106, the server performs neural network model training and detection according to the sample data to obtain a trained neural network model corresponding to the object to be detected.
Based on the scheme, the second principal component analysis is obtained through two times of principal component analysis, and the neural network model training is carried out based on the spectral features corresponding to the second principal component information and the corresponding hyperspectral images as sample data, so that the accuracy of the neural network model training is ensured, the sample data can be reduced, and the training and detection speed is improved.
In some embodiments of the present application, as shown in fig. 2, before step S101, a method provided in embodiments of the present application may further include the following steps:
s201, the server determines to-be-detected objects for building the neural network model based on input information of the user.
For example, if the object to be detected corresponding to the constructed neural network model is an apple, the input information of the user needs to include the apple.
S202, the server obtains an original hyperspectral image of the object to be detected from a prestored database based on the object to be detected and a preset incidence relation.
Corresponding original hyperspectral images can be stored in a preset database in advance, and each original hyperspectral image can carry a label which represents a corresponding object to be detected.
S203, the server performs target detection on each original hyperspectral image to obtain a corresponding region of interest.
As shown in fig. 3, step S203 may be implemented by:
s301, the server determines seed pixels of the original hyperspectral images.
In an embodiment of the application, the server may calculate spectral similarity of pixels in a local window by determining the local window with a preset size, and perform AP clustering to obtain a local window pixel clustering result. And then according to the clustering result, obtaining a local window clustering mark graph. And then, in the clustering label graph, the server searches according to a preset square. And finally, when the gray values of all pixels in the preset grids are equal, the neural network model takes the grid center pixels as seed pixels.
S302, based on the region growing algorithm and the seed pixels, image recognition is carried out on each original hyperspectral image so as to determine target contour information in each original hyperspectral image.
Specifically, the server identifies the image of the original hyperspectral image based on a region production algorithm and the seed pixels to obtain at least one growing region. It is then determined whether each growth region is a connected region. And then the server calculates the area of the connected region and determines a target region according to the area of the connected region so as to obtain corresponding target contour information.
And S303, dividing the corresponding hyperspectral image according to the target contour information to obtain a corresponding region of interest.
In some embodiments of the present application, step 203 may also be implemented by:
the server determines an initial seed pixel of the hyperspectral image, and the spectrum related to the seed is an initial seed spectrum.
It should be noted that determining the initial seed pixel of the hyperspectral image is similar to the method in S301, and is not described herein again.
And the server calculates the spectrum difference between the pixel and the seed spectrum and judges whether the pixel is grown according to the spectrum difference.
And after the image element is grown, the server updates the initial seed spectrum to the spectrum mean value of all grown image elements until the region growth is finished, and the background region is segmented.
And then the server determines a new initial seed pixel, performs region growing again, and performs background region segmentation until the new initial seed pixel cannot be determined so as to obtain a corresponding region of interest.
In an embodiment of the present application, before obtaining the corresponding region of interest, the method further includes:
performing baseline correction on the original hyperspectral image;
according to the following formula:
wherein, the
Is the reflectivity of the food to be measured,
as the reflectivity of the standard reference plate,
to correct the value of the object in the hyperspectral image,
to calibrate the values of the reference plate in the corrected hyperspectral image,
collecting systematic errors for the hyperspectral image;
and carrying out radiometric calibration on the corrected hyperspectral image according to the calibration coefficient to obtain the hyperspectral image after auxiliary calibration.
And S204, the server divides the corresponding original hyperspectral image according to the region of interest to obtain a region of interest image.
And the images of the interested areas form a hyperspectral image set of the object to be detected.
In some embodiments of the present application, after step S106, the method provided in embodiments of the present application may further include the following steps:
and receiving an initial hyperspectral image sent by hyperspectral image acquisition equipment.
And carrying out normalization processing on the initial hyperspectral image to obtain a hyperspectral image to be detected.
And based on the trained neural network model, carrying out image recognition on the hyperspectral image to be detected, and determining whether the initial hyperspectral image comprises the object to be detected.
Through the scheme, the hyperspectral image can be identified, and whether the target object in the hyperspectral image is the object to be detected or not is determined.
Based on the same idea, some embodiments of the present application further provide a device and a non-volatile computer storage medium corresponding to the above method.
Fig. 4 is a schematic structural diagram of a hyperspectral data processing device according to an embodiment of the application. As shown in fig. 4, the apparatus includes:
at least one processor; and
a memory communicatively coupled to the at least one processor;
wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to:
acquiring a hyperspectral image set of an object to be detected;
performing principal component analysis on each hyperspectral image in the hyperspectral image set according to a principal component analysis method to obtain first principal component information of a first preset number;
determining a corresponding characteristic waveband according to a weight coefficient curve in the first principal component system information;
performing principal component analysis on the characteristic wave band according to a principal component analysis method, and determining second principal component information with a second preset number;
taking the spectral characteristics corresponding to the second principal component information and the corresponding hyperspectral image as sample data;
and training and detecting a neural network model according to the sample data to obtain a trained neural network model corresponding to the object to be detected.
Some embodiments of the present application provide a non-volatile computer storage medium corresponding to one of the hyperspectral data processing of fig. 1, storing computer-executable instructions configured to:
acquiring a hyperspectral image set of an object to be detected;
performing principal component analysis on each hyperspectral image in the hyperspectral image set according to a principal component analysis method to obtain first principal component information of a first preset number;
determining a corresponding characteristic waveband according to a weight coefficient curve in the first principal component system information;
performing principal component analysis on the characteristic wave band according to a principal component analysis method, and determining second principal component information with a second preset number;
taking the spectral characteristics corresponding to the second principal component information and the corresponding hyperspectral image as sample data;
and training and detecting a neural network model according to the sample data to obtain a trained neural network model corresponding to the object to be detected.
The embodiments in the present application are described in a progressive manner, and the same and similar parts among the embodiments can be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the device and media embodiments, the description is relatively simple as it is substantially similar to the method embodiments, and reference may be made to some descriptions of the method embodiments for relevant points.
The device and the medium provided by the embodiment of the application correspond to the method one to one, so the device and the medium also have the similar beneficial technical effects as the corresponding method, and the beneficial technical effects of the method are explained in detail above, so the beneficial technical effects of the device and the medium are not repeated herein.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.
Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The above description is only an example of the present application and is not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.
Claims (10)
1. A hyperspectral data processing method, characterized in that the method comprises:
acquiring a hyperspectral image set of an object to be detected;
performing principal component analysis on each hyperspectral image in the hyperspectral image set according to a principal component analysis method to obtain first principal component information of a first preset number;
determining a corresponding characteristic waveband according to a weight coefficient curve in the first principal component system information;
performing principal component analysis on the characteristic wave band according to a principal component analysis method, and determining second principal component information with a second preset number;
taking the spectral characteristics corresponding to the second principal component information and the corresponding hyperspectral image as sample data;
and training and detecting a neural network model according to the sample data to obtain a trained neural network model corresponding to the object to be detected.
2. The method according to claim 1, characterized in that before acquiring the set of hyperspectral images of the object to be detected, the method further comprises:
determining an object to be detected for constructing a neural network model based on input information of a user;
acquiring an original hyperspectral image of the object to be detected from a prestored database based on the object to be detected and a preset incidence relation;
carrying out target detection on each original hyperspectral image to obtain a corresponding region of interest;
segmenting the corresponding original hyperspectral image according to the region of interest to obtain a region of interest image;
and the images of the interested areas form a hyperspectral image set of the object to be detected.
3. The method of claim 1, wherein after obtaining the trained neural network model corresponding to the object to be detected, the method further comprises:
receiving an initial hyperspectral image sent by hyperspectral image acquisition equipment;
carrying out normalization processing on the initial hyperspectral image to obtain a hyperspectral image to be detected;
and carrying out image recognition on the hyperspectral image to be detected based on the trained neural network model, and determining whether the initial hyperspectral image comprises the object to be detected.
4. The method according to claim 2, wherein the performing of the target detection on each of the raw hyperspectral images to obtain a corresponding region of interest specifically comprises:
determining seed pixels of each original hyperspectral image;
based on a region growing algorithm and the seed pixels, carrying out image recognition on each original hyperspectral image to determine target contour information in each original hyperspectral image;
and dividing the corresponding hyperspectral image according to the target contour information to obtain a corresponding region-of-interest image.
5. The method according to claim 4, wherein the determining of the seed pel of each of the original hyperspectral images specifically comprises:
determining a local window with a preset size, calculating the spectral similarity of pixels in the local window, and performing AP clustering to obtain a local window pixel clustering result;
obtaining a local window clustering mark graph according to the clustering result;
searching according to a preset square grid in the clustering label graph;
and when the gray values of all pixels in the preset square are equal, taking the central pixel of the square as a seed pixel.
6. The method according to claim 4, wherein based on a region growing algorithm and the seed pixels, performing image recognition on each of the original hyperspectral images to determine target contour information in each of the original hyperspectral images, specifically comprising:
based on a region production algorithm and the seed pixels, carrying out image recognition on the original hyperspectral image to obtain at least one growing region;
determining whether each of the growth regions is a connected region;
and calculating the area of the connected region, and determining the target region according to the area of the connected region to obtain corresponding target contour information.
7. The method according to claim 2, wherein the performing of the target detection on each of the raw hyperspectral images to obtain a corresponding region of interest specifically comprises:
determining an initial seed pixel of a hyperspectral image, wherein a spectrum related to the seed is an initial seed spectrum;
calculating the spectrum difference between the pixel and the seed spectrum, and judging whether the pixel is grown according to the spectrum difference;
after the pixels are grown, updating the initial seed spectrum to the spectrum mean value of all grown pixels until the region growth is finished, and segmenting the background region;
and determining a new initial seed pixel, re-growing the region, and segmenting the background region until the new initial seed pixel cannot be determined so as to obtain a corresponding region of interest.
8. The method of claim 2, wherein prior to obtaining the respective region of interest, the method further comprises:
performing baseline correction on the original hyperspectral image; and
calculating a calibration coefficient based on the reflectivity of the object to be detected and the reflectivity of the standard reference plate;
and carrying out radiometric calibration on the corrected hyperspectral image according to the calibration coefficient to obtain the hyperspectral image after auxiliary calibration.
9. A hyperspectral data processing apparatus, the apparatus comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor;
wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to:
acquiring a hyperspectral image set of an object to be detected;
performing principal component analysis on each hyperspectral image in the hyperspectral image set according to a principal component analysis method to obtain first principal component information of a first preset number;
determining a corresponding characteristic waveband according to a weight coefficient curve in the first principal component system information;
performing principal component analysis on the characteristic wave band according to a principal component analysis method, and determining second principal component information with a second preset number;
taking the spectral characteristics corresponding to the second principal component information and the corresponding hyperspectral image as sample data;
and training and detecting a neural network model according to the sample data to obtain a trained neural network model corresponding to the object to be detected.
10. A non-transitory computer storage medium for hyperspectral data processing, storing computer-executable instructions, the computer-executable instructions configured to:
acquiring a hyperspectral image set of an object to be detected;
performing principal component analysis on each hyperspectral image in the hyperspectral image set according to a principal component analysis method to obtain first principal component information of a first preset number;
determining a corresponding characteristic waveband according to a weight coefficient curve in the first principal component system information;
performing principal component analysis on the characteristic wave band according to a principal component analysis method, and determining second principal component information with a second preset number;
taking the spectral characteristics corresponding to the second principal component information and the corresponding hyperspectral image as sample data;
and training and detecting a neural network model according to the sample data to obtain a trained neural network model corresponding to the object to be detected.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110453331.0A CN113160183B (en) | 2021-04-26 | 2021-04-26 | Hyperspectral data processing method, device and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110453331.0A CN113160183B (en) | 2021-04-26 | 2021-04-26 | Hyperspectral data processing method, device and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113160183A true CN113160183A (en) | 2021-07-23 |
| CN113160183B CN113160183B (en) | 2022-06-17 |
Family
ID=76870873
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110453331.0A Active CN113160183B (en) | 2021-04-26 | 2021-04-26 | Hyperspectral data processing method, device and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113160183B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113655072A (en) * | 2021-08-20 | 2021-11-16 | 福建中烟工业有限责任公司 | Method, apparatus and computer readable medium for detecting contaminants on a surface of a sample |
| CN113933248A (en) * | 2021-10-12 | 2022-01-14 | 广东省科学院智能制造研究所 | Hyperspectral online detection system and method |
| CN116448689A (en) * | 2023-06-16 | 2023-07-18 | 天津博霆光电技术有限公司 | Single-light-frequency object full-structure detection method and device based on pixel differentiation |
| CN117405607A (en) * | 2023-12-15 | 2024-01-16 | 慧诺云谱(海南)科技有限公司 | Seed phenotype measurement system, method and related equipment |
| CN117557917A (en) * | 2024-01-11 | 2024-02-13 | 杭州海康威视数字技术股份有限公司 | Water quality detection method and device |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101980294A (en) * | 2010-09-25 | 2011-02-23 | 西北工业大学 | Remote sensing image-based method for detecting ice flood of Yellow River |
| CN103279945A (en) * | 2013-04-26 | 2013-09-04 | 北京理工大学 | Quality map guide method and branch cutting method-based interferogram unwrapping method |
| CN104598886A (en) * | 2015-01-23 | 2015-05-06 | 中国矿业大学(北京) | Method for identifying moldy peanuts by using near-infrared high-spectrum image |
| CN106446875A (en) * | 2016-11-16 | 2017-02-22 | 航天恒星科技有限公司 | County remote-sensing scale-oriented crop planting area information extraction method and device |
| CN107274420A (en) * | 2017-06-15 | 2017-10-20 | 中国水产科学研究院东海水产研究所 | The oceanic front extracting method split based on image |
| CN107330875A (en) * | 2017-05-31 | 2017-11-07 | 河海大学 | Based on the forward and reverse heterogeneous water body surrounding enviroment change detecting method of remote sensing images |
| CN108428220A (en) * | 2018-03-05 | 2018-08-21 | 武汉大学 | Satellite sequence remote sensing image sea island reef region automatic geometric correction method |
| CN109308688A (en) * | 2018-09-25 | 2019-02-05 | 中国农业科学院农业资源与农业区划研究所 | A kind of visible light and near infrared band is spissatus and shadow removal method |
| CN109389608A (en) * | 2018-10-19 | 2019-02-26 | 山东大学 | There is the fuzzy clustering image partition method of noise immunity using plane as cluster centre |
| CN110147773A (en) * | 2019-05-23 | 2019-08-20 | 广东工业大学 | A kind of remote sensing images recognition methods |
| CN110852227A (en) * | 2019-11-04 | 2020-02-28 | 中国科学院遥感与数字地球研究所 | Hyperspectral image deep learning classification method, device, equipment and storage medium |
| CN110853062A (en) * | 2019-11-20 | 2020-02-28 | 中国科学院遥感与数字地球研究所 | Rapid segmentation method of multispectral optical remote sensing image |
| CN112070008A (en) * | 2020-09-09 | 2020-12-11 | 武汉轻工大学 | Hyperspectral image feature identification method, device and equipment and storage medium |
| CN112183426A (en) * | 2020-10-10 | 2021-01-05 | 中国科学院空天信息创新研究院 | Method for determining target area in hyperspectral image and related device |
| CN112257496A (en) * | 2020-09-14 | 2021-01-22 | 中国电力科学研究院有限公司 | Deep learning-based power transmission channel surrounding environment classification method and system |
| CN112633045A (en) * | 2019-10-09 | 2021-04-09 | 华为技术有限公司 | Obstacle detection method, device, equipment and medium |
-
2021
- 2021-04-26 CN CN202110453331.0A patent/CN113160183B/en active Active
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101980294A (en) * | 2010-09-25 | 2011-02-23 | 西北工业大学 | Remote sensing image-based method for detecting ice flood of Yellow River |
| CN103279945A (en) * | 2013-04-26 | 2013-09-04 | 北京理工大学 | Quality map guide method and branch cutting method-based interferogram unwrapping method |
| CN104598886A (en) * | 2015-01-23 | 2015-05-06 | 中国矿业大学(北京) | Method for identifying moldy peanuts by using near-infrared high-spectrum image |
| CN106446875A (en) * | 2016-11-16 | 2017-02-22 | 航天恒星科技有限公司 | County remote-sensing scale-oriented crop planting area information extraction method and device |
| CN107330875A (en) * | 2017-05-31 | 2017-11-07 | 河海大学 | Based on the forward and reverse heterogeneous water body surrounding enviroment change detecting method of remote sensing images |
| CN107274420A (en) * | 2017-06-15 | 2017-10-20 | 中国水产科学研究院东海水产研究所 | The oceanic front extracting method split based on image |
| CN108428220A (en) * | 2018-03-05 | 2018-08-21 | 武汉大学 | Satellite sequence remote sensing image sea island reef region automatic geometric correction method |
| CN109308688A (en) * | 2018-09-25 | 2019-02-05 | 中国农业科学院农业资源与农业区划研究所 | A kind of visible light and near infrared band is spissatus and shadow removal method |
| CN109389608A (en) * | 2018-10-19 | 2019-02-26 | 山东大学 | There is the fuzzy clustering image partition method of noise immunity using plane as cluster centre |
| CN110147773A (en) * | 2019-05-23 | 2019-08-20 | 广东工业大学 | A kind of remote sensing images recognition methods |
| CN112633045A (en) * | 2019-10-09 | 2021-04-09 | 华为技术有限公司 | Obstacle detection method, device, equipment and medium |
| CN110852227A (en) * | 2019-11-04 | 2020-02-28 | 中国科学院遥感与数字地球研究所 | Hyperspectral image deep learning classification method, device, equipment and storage medium |
| CN110853062A (en) * | 2019-11-20 | 2020-02-28 | 中国科学院遥感与数字地球研究所 | Rapid segmentation method of multispectral optical remote sensing image |
| CN112070008A (en) * | 2020-09-09 | 2020-12-11 | 武汉轻工大学 | Hyperspectral image feature identification method, device and equipment and storage medium |
| CN112257496A (en) * | 2020-09-14 | 2021-01-22 | 中国电力科学研究院有限公司 | Deep learning-based power transmission channel surrounding environment classification method and system |
| CN112183426A (en) * | 2020-10-10 | 2021-01-05 | 中国科学院空天信息创新研究院 | Method for determining target area in hyperspectral image and related device |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113655072A (en) * | 2021-08-20 | 2021-11-16 | 福建中烟工业有限责任公司 | Method, apparatus and computer readable medium for detecting contaminants on a surface of a sample |
| CN113655072B (en) * | 2021-08-20 | 2023-12-26 | 福建中烟工业有限责任公司 | Method, apparatus and computer readable medium for detecting contaminants on a sample surface |
| CN113933248A (en) * | 2021-10-12 | 2022-01-14 | 广东省科学院智能制造研究所 | Hyperspectral online detection system and method |
| CN116448689A (en) * | 2023-06-16 | 2023-07-18 | 天津博霆光电技术有限公司 | Single-light-frequency object full-structure detection method and device based on pixel differentiation |
| CN116448689B (en) * | 2023-06-16 | 2023-08-29 | 天津博霆光电技术有限公司 | Single-light-frequency object full-structure detection method and device based on pixel differentiation |
| CN117405607A (en) * | 2023-12-15 | 2024-01-16 | 慧诺云谱(海南)科技有限公司 | Seed phenotype measurement system, method and related equipment |
| CN117405607B (en) * | 2023-12-15 | 2024-02-23 | 慧诺云谱(海南)科技有限公司 | Seed phenotype measurement system, method and related equipment |
| CN117557917A (en) * | 2024-01-11 | 2024-02-13 | 杭州海康威视数字技术股份有限公司 | Water quality detection method and device |
| CN117557917B (en) * | 2024-01-11 | 2024-05-03 | 杭州海康威视数字技术股份有限公司 | Water quality detection method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113160183B (en) | 2022-06-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113160183B (en) | Hyperspectral data processing method, device and medium | |
| CN111738165B (en) | Method for extracting individual plant canopy from high-resolution unmanned aerial vehicle visible light remote sensing image | |
| US9230168B2 (en) | Automatic generation of built-up layers from high resolution satellite image data | |
| CN108428220B (en) | Automatic geometric correction method for ocean island reef area of remote sensing image of geostationary orbit satellite sequence | |
| CN115272860B (en) | Determination method and system for rice planting area, electronic equipment and storage medium | |
| CN112053371A (en) | Water body extraction method and device in remote sensing image | |
| CN107316296B (en) | Remote sensing image change detection method and device based on logarithmic transformation | |
| CN112974303A (en) | Hyperspectrum-based fruit quality detection method, device and medium | |
| CN115953371A (en) | Insulator defect detection method, device, equipment and storage medium | |
| CN113516059B (en) | Solid waste identification method and device, electronic device and storage medium | |
| CN117392564B (en) | River water quality inversion method based on deep learning, electronic equipment and storage medium | |
| CN114220017A (en) | Remote sensing data scale self-adaptive adjusting method and device, storage medium and equipment | |
| CN114240940A (en) | Cloud and cloud shadow detection method and device based on remote sensing image | |
| CN112949657B (en) | Forest land distribution extraction method and device based on remote sensing image texture features | |
| CN117710833A (en) | Mapping geographic information data acquisition method and related device based on cloud computing | |
| Agmalaro et al. | Sentinel 1 classification for garlic land identification using support vector machine | |
| CN117423009A (en) | Method and device for determining change condition of ecological system and electronic equipment | |
| CN116977861A (en) | Remote sensing identification method and system for coal-covered area | |
| CN116310876A (en) | Multi-classification identification method for remote sensing image of field structure | |
| CN115719453A (en) | Rice planting structure remote sensing extraction method based on deep learning | |
| CN113096048B (en) | Method and system for radiation correction of generalized cloud-driven and geometric collaborative remote sensing image | |
| Berveglieri et al. | Forest cover change analysis based on temporal gradients of the vertical structure and density | |
| CN113537131A (en) | Land resource analysis model training method and analysis method based on image recognition | |
| CN118655098B (en) | Water quality parameter classification inversion method and system based on spectral reflection | |
| CN118505786A (en) | Remote sensing acquisition method and device for irregular area of crop planting |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: A Hyperspectral Data Processing Method, Equipment, and Medium Effective date of registration: 20230519 Granted publication date: 20220617 Pledgee: Bank of Beijing Co.,Ltd. Jinan Branch Pledgor: Shandong Shenlan Zhipu Digital Technology Co.,Ltd. Registration number: Y2023980041054 |
|
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PC01 | Cancellation of the registration of the contract for pledge of patent right |
Granted publication date: 20220617 Pledgee: Bank of Beijing Co.,Ltd. Jinan Branch Pledgor: Shandong Shenlan Zhipu Digital Technology Co.,Ltd. Registration number: Y2023980041054 |
|
| PC01 | Cancellation of the registration of the contract for pledge of patent right | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: A hyperspectral data processing method, equipment, and medium Granted publication date: 20220617 Pledgee: Huaxia Bank Co.,Ltd. Jinan Branch Pledgor: Shandong Shenlan Zhipu Digital Technology Co.,Ltd. Registration number: Y2024980008356 |
|
| PE01 | Entry into force of the registration of the contract for pledge of patent right |